def test_find_stress(self):
test_string = "reflect respect recline reduce obsessively demonstrate baseball cloud brother cobblestone " +\
"complete conspire estuary"
raw_cmu = transcribe.get_cmu(test_string.split(" "))
result = []
for word_list in raw_cmu:
for word in word_list:
stressed = stress.find_stress(word)
result.append(stressed)
self.assertEqual(result, [
'r ah ˈf l eh k t', 'r ih ˈf l eh k t', 'r ih ˈs p eh k t',
'r iy ˈs p eh k t', 'r ih ˈk l ay n', 'r ih ˈd uw s',
'aa b ˈs eh s ih v l iy', 'ˈd eh m ah n ˌs t r ey t',
'ˈb ey s ˈb ao l', 'k l aw d', 'ˈb r ah dh er',
'ˈk aa b ah l ˌs t ow n', 'k ah m ˈp l iy t', 'k ah n ˈs p ay er',
'ˈeh s ch uw ˌeh r iy'
])
# test the retrieval of only primary stress
self.assertEqual(
stress.find_stress("d eh1 m ah0 n s t r ey2 t", type="primary"),
'ˈd eh m ah n s t r ey t')
# test the retrieval of only secondary stress
self.assertEqual(
stress.find_stress("d eh1 m ah0 n s t r ey2 t", type="secondary"),
'd eh m ah n ˌs t r ey t')
def get_rhymes(word):
if len(word.split()) > 1:
return [get_rhymes(w) for w in word.split()]
phones = remove_onset(preprocess(word))
phones_full = get_cmu([preprocess(word)])[0][0]
c.execute(
f"SELECT word, phonemes FROM dictionary WHERE phonemes "
f"LIKE \"%{phones}\" AND NOT word=\"{word}\" " # don't count word as its own rhyme
f"AND NOT phonemes=\"{phones_full}\""
) # don't return results that are the same but spelled differently
return sorted(list(set([r[0] for r in c.fetchall()])))
def get_rhymes(word, mode="sql"):
if len(word.split()) > 1:
return [get_rhymes(w) for w in word.split()]
phones = remove_onset(preprocess(word))
phones_full = get_cmu([preprocess(word)])[0][0]
if mode == "sql":
c = mode_type(mode)
c.execute(f"SELECT word, phonemes FROM dictionary WHERE phonemes "
f"LIKE \"%{phones}\" AND NOT word=\"{word}\" " # don't count word as its own rhyme
f"AND NOT phonemes=\"{phones_full}\"")
# also don't return results that are the same but spelled differently
return sorted(list(set([r[0] for r in c.fetchall()])))
elif mode == "json":
r_list = []
for key, val in mode_type(mode).items():
for v in val:
if v.endswith(phones) and word != key and v != phones_full:
r_list.append(key)
return sorted(set(r_list))
def get_rhymes(word, mode="sql"):
if len(word.split()) > 1:
return [get_rhymes(w) for w in word.split()]
phones = remove_onset(preprocess(word))
phones_full = get_cmu([preprocess(word)])[0][0]
asset = ModeType(mode=mode).mode
if mode == "sql":
asset.execute(
"SELECT word, phonemes FROM dictionary WHERE phonemes "
"LIKE \"%{0}\" AND NOT word=\"{1}\" ".format(phones, word) +
"AND NOT phonemes=\"{0}\"".format(phones_full))
# also don't return results that are the same but spelled differently
return sorted(list(set([r[0] for r in asset.fetchall()])))
elif mode == "json":
r_list = []
for key, val in asset.items():
for v in val:
if v.endswith(phones) and word != key and v != phones_full:
r_list.append(key)
return sorted(set(r_list))
def syllable_count(word: str, db_type="sql"):
"""transcribes a regular word to CMU to fetch syllable count"""
if len(word.split()) > 1:
return [syllable_count(w) for w in word.split()]
word = transcribe.get_cmu([transcribe.preprocess(word)], db_type=db_type)
return cmu_syllable_count(word[0][0])
def remove_onset(word_in):
phone_list = get_cmu([word_in])[0][0].split(" ")
for i, phoneme in enumerate(phone_list):
if "1" in phoneme:
return ' '.join(phone_list[i:])